Voicemail is not a safety net. It is a record of the calls that were already lost.
The question is not how many people leave messages. It is what you can conclude about the ones who did not.
Most businesses treat voicemail as a backstop: if the call matters, they will leave a message. The logic fails in one specific and expensive way. A voicemail is worth leaving only if you have already decided you want this business in particular, which describes existing customers well and first-time callers comparing options badly. The people most likely to skip the message are the ones who have not chosen you yet, and they are also the ones a business most needs.
Who actually leaves a message?
People with a reason to wait. An existing customer chasing something specific, someone who has already committed, or a caller with no alternative. A stranger with three tabs open and two more numbers to try has none of those reasons.
Our own data gives a sense of what is at stake in that asymmetry. Across a full fifteen-week season, 86% of the calls to one landscaping client came from people who had never called the business before, 69 of 80. If a business with that composition relies on voicemail, it is relying on it precisely where it works worst.
The reverse case is a business whose callers are mostly existing customers. There, voicemail catches a much larger share, and the missing-caller problem is genuinely smaller. Which is the useful point: whether voicemail is a reasonable backstop depends on who calls you, and most businesses have never checked.
What does an empty voicemail box actually tell you?
Less than it appears to. An empty box is consistent with nobody having called, and equally consistent with everyone who called deciding a message was not worth leaving. Without the underlying call log, the two are indistinguishable, and they imply opposite things.
This is the same measurement problem that makes missed calls invisible generally. The absence of evidence is being read as evidence of absence, and the data required to tell them apart is usually one report away in a phone provider's dashboard.
The practical test is simple: compare unanswered calls to messages left, over the same period. The ratio between them is the size of the assumption a business is making when it treats voicemail as a safety net.
What works better than voicemail?
Almost anything that responds while the caller is still present. Answering is best. A text-back within seconds is a distant second and still far ahead of a message nobody left. The ordering follows from the fact that attention has already been paid: the caller is there now.
If voicemail is going to remain the fallback, the useful improvements are unglamorous: keep the greeting short, say when someone will call back, and make sure the box is actually monitored. A greeting that runs forty seconds before the beep is testing the patience of the one caller who was willing to wait.
One caution on text-backs: a missed-call text is a commercial electronic message and sits under CASL. Implied consent from an inquiry runs six months, which usually covers it, but the message still needs identification and an unsubscribe path.
A voicemail is only worth leaving if you have already decided you want this business in particular, which means the callers least likely to leave one are the strangers comparing options, and those are the calls a business most needs.
Compare your unanswered calls to your voicemail count for one month. The gap is the size of the assumption you have been making.
Start free pilotWhat this page does not claim
This is an argument about who leaves messages, supported by one client's caller composition. It is not a general statistic.
- No voicemail abandonment percentage. Figures circulate for how few callers leave messages, and none of them traces back to a stated methodology, so none is quoted here.
- One business's composition. The 86% first-time caller figure covers one landscaping client over one season. It illustrates the mechanism; it does not generalise.
- No claim about our own voicemail data. The lines we build answer rather than divert, so we have no comparative voicemail data to publish.
Questions this raises
- Do people still leave voicemails?
- Some do, mostly people who have already chosen the business. This page does not quote a percentage, because the widely circulated figures could not be traced to a stated methodology.
- Is an empty voicemail box a good sign?
- It is ambiguous. It is equally consistent with nobody calling and with everybody who called deciding not to leave a message. The call log distinguishes them; the voicemail box cannot.
- Which callers are least likely to leave a message?
- First-time callers comparing options. Leaving a message assumes you want this business specifically, which is exactly what a stranger has not decided yet.
- Is a missed-call text better than voicemail?
- Generally yes, because it reaches the caller while the need is still live. It is also a commercial electronic message under CASL, so it needs identification and an unsubscribe path.
- How do I tell whether voicemail is working for my business?
- Compare unanswered calls to messages left over the same period. If the gap is large, voicemail is not the backstop it is being treated as.
Sources
- Canada's Anti-Spam Legislation: Guidance on Implied ConsentCRTCconsulted 2026-08-09
- Lead Response Management Study (Dr. James Oldroyd)MIT / InsideSales.com2007, consulted 2026-08-09
Want an AI receptionist built around these rules rather than despite them?
We build a working receptionist on your calls, your booking rules and your escalation rules, and you call it yourself before committing to anything. About 14 days, no call limit during the pilot.
Start free pilot